Combining ability and heterosis for stem sugar traits and grain yield components in dual-purpose sorghum (Sorghum bicolor L. Moench) germplasm.

(1)

Combining Ability and Heterosis for Stem Sugar Traits and Grain Yield Components in Dual-Purpose Sorghum

**(Sorghum bicolor L. Moench) Germplasm**

By

Itai Makanda

MSc. Crop Science (Plant Breeding) (University of Zimbabwe) BSc. Agric. Hons. Crop Science (University of Zimbabwe)

A thesis submitted in partial fulfilment of the requirements for the degree of Doctor of Philosophy (PhD) in Plant Breeding

African Centre for Crop Improvement School of Agricultural Sciences and Agribusiness

Faculty of Science and Agriculture University of KwaZulu-Natal

Republic of South Africa December 2009

(2)

ii

Abstract

Sorghum is the fifth most important cereal crop in the world and ranks third in Africa, and it is potentially the number one cereal for the semi-arid environments in sub-Saharan Africa.

Sorghum varieties have been developed specifically for grain, fodder or stem sugar but not for dual-purpose combining grain and stem sugar. Such varieties could be beneficial to the resource-poor farmers by providing grain for food and sugar rich stalks that can be sold for bioethanol production. However, there are no suitable dual-purpose cultivars on the market.

There is also limited information about the combining ability, gene action and genetic effects and relationships between stem sugar and grain yield which is required in devising appropriate strategies for developing dual-purpose sorghum varieties. Furthermore, there is also lack of information about the perceptions of resource-poor, small-scale farmers and other important stakeholders on the potential of dual-purpose sorghum production and the value chain.

Therefore, the objectives of this study were to: (i) investigate the awareness of the farmers, industry and other stakeholders on the dual-purpose sorghum varietal development and its feasibility, (ii) screen germplasm for use as source materials useful for grain yield and stem sugar traits, (iii) investigate the inheritance and heterosis levels attainable in grain yield components and stem sugar traits in dual-purpose sorghums, (iv) determine the relationships between stem sugar traits and grain yield components in dual-purpose sorghums, and (v) investigate the fertility restoration capacities of selected male-fertile lines used as male parents through the evaluating seed set in experimental dual-purpose hybrids.

Two surveys were conducted to establish stakeholders’ level of awareness and perceptions on the potential and feasibility of developing and utilising dual-purpose sorghums in Southern Africa.

One survey was carried out in the semi-arid tropical lowlands in Zimbabwe under the conditions of small-scale and resource-poor farmers while the other, which targeted sugar industries, plant breeders, engineers, political leaders, economists and extension workers, was conducted in South Africa and Zimbabwe. Data were analysed using SPSS computer package. Results showed that both farmers and the non-farmer stakeholders were in agreement on the view that dual-purpose sorghum would be a viable enterprise that could alleviate poverty, enhance food security, create rural employment and boost rural development in southern African countries.

Farmers were willing to adopt the cultivars if they were made available. The stakeholders also suggested mechanisms to overcome the infrastructural, economic and technical challenges associated with the technology.

Screening of regional and international germplasm collection held at the University of KwaZulu- Natal in South Africa revealed high genetic variability for grain yield, stem brix and stem biomass yield that can be exploited in dual-purpose sorghum cultivar development. Ten lines were selected for inclusion as parents in the dual-purpose sorghum breeding programme. The selections were crossed to eight cytoplasmic male-sterile lines originating from the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) in accordance with a North Carolina Design II mating scheme. The 18 parents, together with the 80 experimental hybrids

(3)

iii generated and two check varieties were evaluated for grain yield and stem sugar traits in six tropical low- and mid-altitude environments in Mozambique, South Africa and Zimbabwe. Stem sugar concentration and stem biomass were measured at the hard dough stage of each entry due to maturity differences between the genotypes. Grain yield was measured and adjusted to 12.5% moisture content. Data were analysed in GenStat computer package following a fixed effects model. Both additive and non-additive gene effects were important in controlling stem brix, stem biomass, grain yield and the associated traits in dual-purpose sorghum. This showed that breeding progress can be achieved through hybridisation and selection. Cultivars showing high stability, and high standard and better-parent heterosis for the three traits were identified implying that breeding for general adaptation was an option and that productivity could be enhanced by breeding hybrid cultivars.

The relationships between traits were estimated using correlation and path-coefficients analysis.

Grain yield was found to be negatively and significantly associated with stem brix but was positively and significantly associated with stem biomass. This implied that breeding for high stem brix might compromise grain yield but selection for high stem biomass improved grain yield.

Stem biomass and stem brix were not significantly correlated. The general negative relationship between grain and stem brix was attributed to the predominance of entries with contrasting performances for the two traits. However, the relationship between grain yield and stem brix of the top 20 performing entries showed a non-significant relationship between stem brix and grain yield suggesting that the traits were independent of each other. This finding was confirmed by the presence of crosses that combined high performance for both stem brix and grain yield as well as stem biomass among the hybrids. The relationships between stem brix and stem biomass for the top 20 performers remained non-significant while that between stem biomass and grain yield became stronger, positive and significant. Direct selection for stem brix and grain yield was shown to be more important than indirect selection, while selection for stem biomass improves grain yield but had no effect on stem brix. Therefore, it is possible to breed dual- purpose sorghum cultivars and the identification of genotypes combining the desirable traits is prudent in addition to general relationships information.

The study on fertility restoration capacities as evaluated through hybrid seed set showed that fertility restoration was under the control of genes with both additive and non-additive action.

Since restoration is conferred by a single dominant gene (Rf1), this could have arisen from the action of the modifier genes that have been previously reported to influence it. This showed that fertility restoration can be improved through breeding. Hybrid combinations showing complete seed set and high performance for grain, stem brix and stem biomass were identified and are potential dual-purpose sorghum cultivars. Overall, the study showed that development of dual- purpose sorghum cultivars would be feasible and genotypes identified as potential cultivars in this study will be forwarded for further testing across many sites and seasons in the target environments.

(4)

iv

Declaration

I, Itai Makanda, declare that

1. The research reported in this thesis, except where otherwise indicated, is my original research.

2. This thesis has not been submitted for any degree or examination at any other university.

3. This thesis does not contain other persons’ data, pictures, graphs or other information, unless specifically acknowledged as being sourced from other persons.

4. This thesis does not contain other persons' writing, unless specifically acknowledged as being sourced from other researchers. Where other written sources have been quoted, then:

a. Their words have been re-written but the general information attributed to them has been referenced

b. Where their exact words have been used, then their writing has been placed in italics and inside quotation marks, and referenced.

5. This thesis does not contain text, graphics or tables copied and pasted from the Internet, unless specifically acknowledged, and the source being detailed in the thesis and in the references sections.

Signed

………..

Itai Makanda

As the candidate’s supervisors, we agree to the submission of this thesis:

………

Prof. Pangirayi Tongoona (Supervisor)

………..………

Dr. John Derera (Co-Supervisor)

(5)

v

Acknowledgements

I would like to thank the following people and organisations for their assistance in making this research a success:

· My supervisor Professor Pangirayi Tongoona and my Co-Supervisor Dr. John Derera of the University of KwaZulu-Natal (African Centre for Crop Improvement), South Africa for their guidance and advice that made this research work successful;

· All the members of academic and administrative staff at the Africa Centre for Crop Improvement (ACCI) for their support during my work;

· Mr. Jurie Steyn of the Agricultural Research Council based at Makhathini Research Station (South Africa); Mr. Pedro Fato and Mr. Egas Nhamucho of the Instituto de Investigação Agraria de Moçambique (Mozambique); and Mr. Caleb Souta and Mr.

Walter Chivasa of Seed Co (Zimbabwe); and Mr. Solomuzi P. Dlamini at Ukulinga Research Farm (South Africa) for their assistance in running the sorghum trials. The assistance of Stanlake Kaziboni, Daniel Mulenga and Stan Thamangani in conducting the surveys in Zimbabwe is greatly appreciated;

· The International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) for providing the germplasm, especially the cytoplasmic male-sterile lines that proved invaluable to this work;

· My family Farayi N. Makanda, Ruth Mae Chiyabi, Lucy Tsirizani, Tarisai “Mwenda”

Makanda, Takudzwa “Wanyoni” Makanda, Mafara “Sevhe” Chimutashu, Esther Makanda, Vimbai (Makanda) Kaitano, Revai Makanda and Professor Peter K. Dzvimbo for supporting and believing in me as I pursued my studies;

· My friends in the different ACCI cohorts including Julia and Pedro “Emmanuel”, Jimmy

“Garwe”, Sharmane, Jean-Baptist, and Eric for walking the path with me; and

· The Alliance for a Green Revolution in Africa (AGRA) for making this research possible through funding this work.

(6)

vi

Dedication

Kurangarira vakatungamira, vasina kumirira kuti vandisusukidze mberi, asi kusiri kuda kwavo.

1. Vabereki vangu Henry Chemerayi Makanda naSarah Tsirizani-Makanda.

2. Vasekuru Chari “Chimusoro” Makanda nambuya Norah Zvinorukwa Makanda.

3. Vasekuru Peter Tsirizani nambuya Gladys Tsirizani 4. Vanonditarisa pakufamba, ndinotenda.

(7)

vii

List of Tables

Table 1.3-1: Estimates of heritability in the broad sense (H²) for grain yield and its components in sorghum ... 17 Table 1.4-1: Possible situations arising from using male parents whose restoring ability is not confirmed 23 Table 2.4-1: Farmer and household information for Chivi and Chipinge districts in Zimbabwe ... 45 Table 2.4-2: Crop history over five years (2002-2006) showing the area planted (ha) to each crop per household in Chivi and Chipinge districts in Zimbabwe... 48 Table 2.4-3: The proportion of people who planted various crops at particular times of the year, sourced seed from various sources and marketed their produce to particular markets ... 51 Table 2.4-4: Proportion of farmers who acknowledged the production constraints and how they ranked them ... 56 Table 3.4-1: Percentages of farmers responding to questions on dual-purpose sorghum in three areas studied in Zimbabwe... 69 Table 3.4-2: Percentage of the other stakeholders responding to questions on the use of dual-purpose sorghum for bio-fuel production in Zimbabwe and South Africa (n = 25) ... 70 Table 3.4-3: Five-year production area planted to the three major crops per household and P-values for each year across the three areas studied in Zimbabwe ... 73 Table 3.4-4: Percentage of farmers preferring certain traits in sorghum in the three case study areas ... 74 Table 4.4-1: Mean squares for sorghum traits measured in 80 genotypes at Ukulinga farm, during

2006/2007 season ... 85 Table 4.4-2: Means of traits measured for the top 30 and bottom five stem brix performers at maturity (data sorted by stalk sugar performance at maturity) ... 86 Table 4.4-3: Correlation coefficients between sorghum traits measured in 80 varieties at Ukulinga farm, during 2006/2007 season... 90 Table 4.4-4: Path coefficient analysis table showing the direct and indirect effects of the sorghum trait to stem sugar at maturity ... 90 Table 5.3-1: Name, origin and pedigree of parental sorghum lines used in the study ... 100 Table 5.4-1: Mean squares for stem brix and associated traits of sorghum hybrids across six environments ... 105 Table 5.4-2: Stem brix of selected sorghum hybrids and parents over six environments (genotype by environment mean matrix SED = 0.95) ... 106 Table 5.4-3: Stem biomass (kg ha^-1) performance of selected sorghum hybrids and parents over six environments (genotype by environment mean matrix SED = 5890) ... 108 Table 5.4-4: Stem juice-brix index and other stem sugar traits of selected sorghum hybrids and parents across six environments ... 109 Table 5.4-5: Per se and inter se cultivar superiority indices (Pi) of 18 sorghum parents for stem brix and stem biomass across six environments ... 111

(14)

xiv

Table 5.4-6: GCA effects for the male and female sorghum parents for stem brix and associated traits across six environments ... 113 Table 6.3-1: Name, origin and pedigree of parental sorghum lines used in the study ... 128 Table 6.4-1: Mean squares for grain yield and associated traits of sorghum hybrids across six

environments... 131 Table 6.4-2: Grain yield (kg ha^-1) of selected sorghum hybrids and parents across six environments (genotype by environment mean matrix SED = 396.8) ... 133 Table 6.4-3: Per se and inter se cultivar superiority indices (Pi) of 18 sorghum parents for grain yield across six environments ... 134 Table 6.4-4: Means of traits associated with grain yield for selected sorghum hybrids and parents across six environments ... 135 Table 6.4-5: Off-season and in-season grain yield performance of selected hybrids and parents ... 136 Table 6.4-6: Correlation coefficients between grain yield and its components for sorghum hybrids and parents across six environments ... 137 Table 6.4-7: GCA effects for the sorghum male and female sorghum parents for grain yield and its

components across six environments ... 140 Table 7.4-1: Correlation coefficients between grain yield and stem sugar content with selected agronomic traits in sorghum ... 157 Table 7.4-2: Direct and indirect path coefficients of selected sorghum traits on grain yield across six environments... 158 Table 7.4-3: Direct and indirect path coefficients of selected sorghum traits on stem brix at maturity across six environments ... 159 Table 7.4-4: Stem brix, stem biomass and grain yield performance and cultivar superiority of selected sorghum hybrids and parents across six environments... 160 Table 8.3-1: Name and origin of the lines used in this study... 171 Table 8.3-2: Rating scale for seed set of sorghum hybrids at two sites in South Africa ... 173 Table 8.4-1: Mean squares and significance of sorghum hybrids and parents at Makhathini Research Station and Ukulinga Research farm during the 2008/09 summer rainy season in South Africa ... 174 Table 8.4-2: Seed set percentages of sorghum hybrids at Makhathini Research Station and Ukulinga Research farm during the 2008/09 summer rainy season in South Africa ... 175 Table 8.4-3: Parental GCA effects for sorghum hybrid seed set percentages across two environments.. 176 Table 8.4-4: Performance data over six environments for selected hybrids differing in seed set levels across two sites ... 178

(15)

xv

List of Figures

Figure 1.3-1: The Atago PAL-1 digital hand-held pocket refractometer used to screen sorghum entries for stem sugar accumulation in the study ... 20 Figure 1.4-1: Maintenance of (i) the maintainer (male-fertile) B-line , (ii) its iso-cytoplasmic male-sterile A- line and (iii) F1 hybrid seed production scheme and (iv) the male fertility restoring R-line maintenance in a sorghum breeding programme ... 24 Figure 2.4-1: National crop productivity and production figures for Zimbabwe (Source: FAOSTAT, 2008) 49 Figure 2.4-2: Levels of fertiliser application for the major crops grown by the communities in Chipinge and Chivi districts in Zimbabwe: (a) Maize, (b) sorghum, (c) pearl millet, (d) groundnut, (e) beans, and (f) cotton ... 52 Figure 4.4-1: Histogram showing variation of the traits among 80 sorghum germplasm evaluated at the University of KwaZulu-Natal (South Africa) in the 2006/2007 rainy season: (a) stem brix at anthesis, (b) stem brix at maturity, (c) seed yield, (d) days to anthesis, (e) stem biomass, (f) plant height, (g) stem diameter... 89 Figure 5.3-1: Long term mean (five-year) temperatures for CRS, URF, RARS and MRS [Data source:

Agricultural Research Council-ISCW AgroMet Potchefstroom (2009); Seed Co. Zimbabwe Ltd (2009);

Gaisma (2007)] ... 101 Figure 5.4-1: Twenty-seven sorghum hybrids exhibiting positive better-parent heterosis for mean stem brix across six environments ... 112 Figure 5.4-2: Twenty-five sorghum hybrids displaying positive better-parent heterosis for stem biomass across six environments ... 112 Figure 5.4-3: Sorghum crosses showing positive and significant SCA effects for (a) stem brix across six environments (SE = 0.65; SED = 0.92) and (b) stem juice score (SE = 0.46; SED = 0.64) ... 114 Figure 5.4-4: Thirty-eight sorghum crosses showing positive and significant SCA effects for stem biomass weight across six environments (SE = 162.4; SED = 229.7)... 115 Figure 6.3-1: Long term (five year) mean temperatures for the sites [Data source: Agricultural Research Council-ISCW AgroMet Potchefstroom (2009); Seed Co. Zimbabwe Ltd (2009); Gaisma (2007)] ... 128 Figure 6.4-1: Twenty-seven sorghum hybrids showing positive better-parent heterosis for mean grain yield ... 138 Figure 6.4-2: Thirty-six sorghum hybrids showing positive better-parent heterosis for head length ... 138 Figure 6.4-3: Sorghum crosses showing positive and significant SCA effects for (a) grain yield (SE = 253.5; SED = 358.5) and (b) head length (SE = 0.8; SED = 1.1) across six environments... 139 Figure 7.3-1: Path diagram showing relationships between grain yield or stem brix against stem biomass, grain yield/stem brix, head length, stem juice score, number of leaves per plant, plant height and stem diameter, ... 156 Figure 8.3-1: Mean annual temperature data for Makhathini Research Station and Ukulinga research Farm during the trial [Data source: Agricultural Research Council-ISCW AgroMet Potchefstroom (2009)] ... 172

(16)

xvi

Figure 8.4-1: Crosses showing significant (P≤0.05) SCA effects for sorghum seed set (SE =10.02; SED = 14.18)... 177

(17)

1

Introduction to thesis

Rationale for sorghum improvement

Sorghum (Sorghum bicolour L. Moench) total production ranks fifth among the important cereal crops worldwide and is second after maize in Africa (Chantereau and Nicou, 1994). In the sub- Saharan Africa, it is arguably the most important cereal. Most of the sorghums planted, especially in Africa, are grain sorghum cultivars. The world produces about 63,375,602t of sorghum grain from 46,928,023ha of which 26,065,312t comes from Africa (from 29,499,987ha) and 156,796t comes from southern Africa from an area of 219,090ha (FAO, 2009). Apart from grain, sorghum can be used for sugar production using the sweet stem types that accumulate fermentable sugars (≥8%) in their stems. These sweet sorghums provide an avenue for transforming sorghum into an industrial crop due to their potential use in bioenergy production.

Stem sugar can be used for bioethanol production and the grain for food. This provides rural households with dietary energy as well as the much needed income for other requirements such as education and health from the small pieces of land from which they subsist. However, no effort has been made to combine grain yield and stem sugar in a single cultivar to produce a dual-purpose sorghum variety in southern Africa regardless of the presence of such varieties reported in other regions (FAO, 2002; Reddy et al., 2005). The potential for both grain yield, stem sugar accumulation and stem biomass yields in dual-purpose sorghum cultivars for the region is not known. This is evidenced by the non-availability of the dual-purpose sorghum cultivars on the market. Generating information on the behaviour of the traits after combination is important for a dual-purpose sorghum breeding programme. An ideal dual-purpose sorghum should achieve acceptable grain yields, stem brix and stem biomass. Although there are no set values of these traits, minimum grain yield of 1.5t ha^-1, stem brix of 11ºbrix and stem biomass of 30t ha^-1 can be arbitrarily set to achieve about 3000l ha^-1and food security. Assuming a farmer plants two hectares, 3t of grain and 6000l of ethanol can be produced based on studies by Woods (2000) and Tsuchihashi and Goto (2004).

The potential for generating bioethanol from sweet sorghum has not been quantified in most countries and environments in southern Africa. However, stem sugar values of between 10%

and 25% (10ºbrix and 25ºbrix) have been reported in the literature (Woods, 2000; Reddy et al., 2005; Tsuchihashi and Goto, 2004, 2008). Stem sugar can be processed into jaggery or distilled to produce ethanol (FAO, 2002). Bioethanol yield of 3000l to 7000l ha^-1 have been reported from biomass levels of between 30t to 120t ha^-1 in Zimbabwe (Woods, 2001); Romania (Roman et al., 1998); Italy (Dolciotti et al., 1998); United States of America (Anderson, 2005); China (FAO,

(18)

2 2002); and various European union countries (Claassen et al., 2004). Mean grain yield potentials of between 1.0t and 6.5t ha^-1 have been reported with improved grain sorghums in Zambia, Zimbabwe and Botswana (Obilana et al., 1997). However, most African farmers use unimproved cultivars with productivity of less than 1.0t ha^-1.

The potential of dual-purpose sorghum in rural economy

The potential use of sweet sorghum for bioethanol production has been demonstrated in southern Africa by Woods (2001) working in Zimbabwe. Sweet sorghum can be successfully incorporated into the sugarcane processing system. Countries in southern Africa, for example Malawi, Mozambique, South Africa, Swaziland, Zambia and Zimbabwe, which have viable sugar industries based on sugarcane (Saccharum officinarum L.), the major sugar producing crop, can benefit from dual-purpose sorghums. These countries can exploit the sugar mills by using sweet sorghum stalks thereby maximising output. The dual-purpose sorghums thus complement sugarcane. Sweet sorghum is widely adapted; can be grown under dryland conditions where sugarcane cannot grow and have rapid growth (Reddy and Sanjana, 2003). Further, sorghum can be ratooned thereby obtaining more than one biomass harvest from a single planting, although research has shown diminishing yields and increased disease pressure from a ratooned crop. Overall, provision of dual-purpose sorghums can lead to sustainable rural development, enhanced renewable energy production, higher health standards through cleaner fuels, and improved food security (Woods, 2001). Farmers in most parts of southern Africa already grow sorghum, for example, in Musikavanhu, a dryland communal area in Zimbabwe.

Chivasa et al. (2001) reported that sorghum was grown by 94% of the farmers in the area and occupied 82% of the land. Derera et al. (2006) reported similar findings in the same area.

On the food security aspect, many world bodies express reservations on the use of bioenergy crops due to potential competition for land with food crops, thereby pushing food prices beyond the reach of many. This necessitated the search for alternative non-food security crops for bioenergy production. Guiying et al. (2000) reported that sorghum cultivars with high stem sugar and high grain yield were required for China. This creates the niche for dual-purpose sorghum where farmers can harvest the grain for food and sell the stalks for sugar extraction, reaping twice from the same crop and same piece of land. According to the FAO (Gnansounou et al., 2005) dual-purpose sorghums can give a yearly gross margin of US$1300 ha^-1 compared to only US US$1054 ha^-1 for grain sorghum reported by Hagos et al. (2009). With specialised sweet sorghum, farmers can earn between US$40 to US$97 ha^-1more than for grain sorghum (PSciJourn, 2010). Thus, dual-purpose sorghum development would not only increase income

(19)

3 for the household, but improve food security as well. The challenge that may arise is the development of appropriate technologies and markets for the stems. However, with the projected future fossil fuel shortages and the increasing global call for cleaner environments, adoption of sweet sorghum as a raw material for the production of alternative fuel is inevitable. This will potentially increase the demand for sweet stem sorghum, making the enterprise viable.

In Zimbabwe, the cost of producing a litre of bioethanol from sweet sorghum was reported at US$0.19 compared to the then global prices of ethanol of US$0.30 to $0.35 a litre (Woods, 2000). Roman et al. (1998) reported 2000-3000TOE (tones of oil equivalence) per hectare of sweet sorghum and 6-9TOE of fuel from bagasse. Up to 9000l ha^-1 of ethanol has been reported in China (FAO, 2002) and Greece (Sakellariou-Makrantonakai et al., 2007). Woods (2000) reported the production of 12.6GJ electricity (3.5MWhe at 15% conversion efficiency) from 46t ha^-1 of fresh stem weight was achieved using sweet sorghum cultivars identified in Zimbabwe.

High grain yield potential of 2.0t to 6.0t ha^-1 has been reported for sweet sorghum (FAO, 2002;

Reddy et al., 2005). The ethanol yields of up to 7000l ha^-1 combining with grain yield of up to 6.0t ha^-1 can potentially make the bioethanol industry, using sweet sorghum, viable at the same time contributing towards food security and social sustainability (Zhao et al., 2009). Therefore, introducing dual-purpose sorghum is likely to impact positively on both food security and rural development.

Exploitation of the off-season production

Successful production of dual-purpose sorghum might entail expansion of production to include off-season production in tropical lowlands that have optimum temperatures and water supply throughout the year. These areas include the Zambezi basins covering parts of Zimbabwe, Mozambique, Zambia and Malawi, Makhathini flats in South Africa and Chokwe in Mozambique.

Off-season production in Chokwe and Makhathini flats was demonstrated to give optimum grain yield, stem biomass and stem sugar percentage with dual-purpose sorghum experimental hybrids. Off-season production for grain was reported to significantly contribute to food security in Somalia (Food Security Assessment Unit, 1998) and India (Patil, 2007), while all year round production of sweet sorghum was demonstrated in Indonesia (Tsuchihashi and Goto, 2008).

There is potential to use the off-season in addition to the traditionally used in-season production environments for dual purpose sorghum. The limitation is the non-availability of appropriate dual- purpose sorghum cultivars for this purpose and cultivar development is viewed as key strategy.

Therefore, it is important to generate the information required towards the devising of an

(20)

4 appropriate strategy to develop dual-purpose sorghum cultivars that are also adapted to the off- season production environments.

The need for appropriate breeding source germplasm

Cultivar development relies on the presence of genetic variability for the traits under consideration. This makes it imperative to evaluate germplasm collections for grain yield potential, stem sugar accumulation and stem biomass potential to select the appropriate parents for the breeding programme. The next step is to understand the gene action controlling the traits of interest. This information can be obtained by conducting combining ability studies that entails systematic crossing of the selected parents using appropriate mating designs and subsequent hybrid evaluation. Information on general combining ability (GCA) and specific combining ability (SCA) effects is critical in cultivar development, either through selection or inbreeding then cross breeding (Goyal and Kurmar, 1991; Cruz and Regazzi, 1994; Kenga et al., 2004). Variation due to GCA is attributed to additive genes, and that due to SCA is attributed to non-additive gene action. Lack of this information limits research under African dry-land sorghum growing conditions (Kenga et al., 2004). For dual-purpose sorghums, knowledge of the inheritance of the major traits, grain yield, stem sugar content and stem biomass, is important. Although a lot of work has been done on the inheritance of these traits separately, no work has been reported on the traits combined in dual-purpose sorghums. In the long run, the development of parental lines (base population) with high performance in stem sugar, grain yield potential and stem biomass is key to the dual-purpose sorghum cultivar breeding programme. These can be used as parents in hybrid cultivar development programmes or as pure line varieties where high performance is demonstrated.

Is southern Africa ready for sorghum hybrids?

The promotion of hybrid cultivars is one way of enhancing productivity in southern Africa without expanding the area under production. There is already a high pressure on pressure, in Malawi and Zimbabwe for example, with serious political consequences. Hybrid cultivars have been shown to be more productive than pure line varieties in sorghum (Li and Li, 1998). Many researchers are now advocating for hybrid sorghum deployment in Africa. Haussmann et al.

(1999) concluded that hybrids could boost sorghum production in Kenya. Hybrids can also bring the private sector into the sorghum industry, which can result in the supply of high quality seed and agronomic support services for the farmers. Farmers in southern Africa, especially Zimbabwe and South Africa, have been purchasing maize hybrid seed for years and are familiar with the benefits of hybrid cultivars. These farmers are likely to adopt hybrid sorghum cultivars

(21)

5 easily, given the economic benefits of dual-purpose sorghums. The Malawi, Zambia and Mozambique governments also have seed and other agronomic input distribution programmes called the Input Voucher systems that can be exploited to promote sorghum hybrids production (Mangisoni et al., 2007). Given the foregoing, it is most likely that hybrid sorghum deployment in the region, especially with an added industrial trait such as stem sugar, can increase sorghum production. It is, therefore, necessary to evaluate the levels of heterosis attainable and stability of the traits. However, effective hybrid development programmes in sorghum are dependent on the identification of heterotic fertility restorer (male) and cytoplasmic male-sterile (female) lines.

Production of grain on the hybrids is important in dual-purpose cultivars. There is need to evaluate hybrid fertility from crosses using male-sterile parents because hybrid fertility restoration has been shown to be influenced by the genetic background and the environment in which the restorer genes are operating (Sleper and Poehlman, 2006).

Rationale for studying the inheritance of grain yield and stem sugar in dual-purpose sorghum

Knowledge of the mode of inheritance of the dual-purpose sorghum traits, mainly grain yield, stem sugar and stem biomass, would be useful for developing a viable breeding programme. A survey of the literature did not identify any source of such information. However, in studies involving sweet stem sorghum, genes with partial dominance, additive effects, and both major and minor effects were reported to control stem sugar in sorghum with trait heritability estimates ranging between 40% and 96% (Schlehuber, 1945; Baocheng et al., 1986). Many studies reported significant GCA and SCA effects for grain yield in sorghum with heritability estimates ranging between 10% and 86% (Haussmann et al., 1999; Kenga et al., 2004; Bello et al., 2007).

This information suggests that the two traits are controlled by genes with both additive and non- additive effects and can, therefore, be improved through hybridisation and selection. However, the behaviour of the genes in dual-purpose sorghum cultivars when grain yield and stem sugar are combined in one cultivar is not known. This information is important in formulating a breeding strategy. For example, if it is found that the behaviour of the genes in dual-purpose cultivars is the same as in the specialised grain and sweet sorghums, then hybridisation and selection can be used in cultivar development. The development and use of parents with high GCA and identification of heterotic crosses showing high SCA effects is expected to maximise breeding gain. However, unless this information is generated, it is difficult to formulate such breeding strategies, hence it is crucial to study the gene action for the traits in dual-purpose sorghum cultivars.

(22)

6 Is high grain yield mutually exclusive of high stem sugar and high stem biomass?

Development of dual-purpose sorghum cultivars entails the combination of high grain yield potential, high stem sugar accumulation and high stem biomass potential in one cultivar. This activity is based on the understanding of the relationships between the traits to devise appropriate selection criteria. There are no conclusive reports regarding the relationships between grain yield components and stem sugar traits in the literature surveyed on specialised cultivars. No reports were found on dual-purpose sorghum. This information is crucial in developing dual-purpose sorghum cultivars and its generation is viewed as an important activity.

Guiying et al. (2000) reported a general negative relationship between stem brix and weight of 1000 seed in sorghum. It can be argued that 1000 seed weight alone does not give a clear picture of the relationship because it is not representative of grain yield per plant or per unit area.

In other studies, Ferraris and Charles-Edwards (1986) found stem sugar concentration to increase as a function of growth duration. Although they reported low grain yields in their cultivars, they also found remobilization of stem sugars to the grain to be negligible. Although the high stem sugar after seed initiation and the resultant low yields may seem to suggest a negative association between stem sugar content and grain yield, the negligible remobilisation of stem sugar to grain after grain initiation suggests otherwise. It is important to conduct detailed studies on the relationship because it influences the selection strategy during cultivar development.

Correlation and path coefficient studies have been shown to establish trait relationships in crops (Ofori, 1996; Makanda et al., 2009).

Rationale for stakeholders involvement and situational studies in dual-purpose sorghum cultivar development

Southern Africa is dominated by small-scale and resource-poor farmers living in the semi-arid low and mid-altitude environments, producing about 80% of the total sorghum crop. These areas account for about 35% of the cereal mega environments (Vivek et al., 2005). Until recently, breeding has been conducted by researchers without involving the farmers and usually at research stations which are not necessarily representative of the conditions in the small-holder or resource-poor farmers’ conditions. This meant a complete marginalisation of the farmers from setting of the research agenda to the formulation of the solutions. This approach has been demonstrated not to be effective due to the uniqueness of the small-scale and resource-poor farmers’ situations. Many researchers have reported on the negative consequences of not including farmers in setting up research and policy agenda (Derera et al., 2006). This led to breeders shifting from the traditional approaches of scientist-centred research agenda to the inclusion of the farmers in problem identification and research agenda formulation (Dixon et al.,

(23)

7 2001). Understanding the farmers is an initial step towards the search for an effective and sustainable way to make agricultural research more relevant to them (Kudadjie et al., 2004).

Although cases in dual-purpose sorghum were not found in the literature, cases of different preferences between farmers and researchers are well documented in specialised sorghum cultivars. In Ethiopia, farmers selected only three varieties from the eight that researchers considered the best (Mulatu and Belete, 2001). In the same study, the farmers also proved wrong the notion that existed among researchers that they were not willing to grow short season varieties.

In Malawi, the use of farmer participation in cultivar evaluation has resulted in the selection of high yielding sorghum landraces in farmers fields (Nkongolo et al., 2008). In both cases, the cultivars were well received by the farmers. This demonstrates the breeding success that can be realised if farmers’ views and preferences are taken into consideration during cultivar development.

Situational studies are very important as a first step in new cultivar development. They generate information about the farmer and their socio-economic conditions that impact on cultivar adoption.

Important information should be established about the levels of knowledge, age, labour, land holding, resource availability, constraints to production, possible competing cropping enterprises and access to produce markets. If not clearly identified, the factors can impact negatively on the dual-purpose sorghum adoption and production. In situations where the farmers and other stakeholders are not familiar with the technology, as is the case with dual-purpose sorghum cultivars in the lowland areas of Zimbabwe, interacting and discussing with the farmers also helps to create awareness. This information can be gathered using participatory research techniques used to gather information prior to, during and after technology deployment (Matata et al., 2001).

These techniques give farmers an avenue for participation in decision-making, especially on the type of cultivars they prefer in the case of dual-purpose sorghums. This increases cultivar adoption rates as was the case in Ethiopia and Malawi. The situational studies can also help to explain the anticipated adoption pattern, which aid future breeding projects for the farmers.

Given the foregoing, it is prudent to include all stakeholders, from farmers to industrial end users, in dual-purpose sorghum cultivars development if acceptable cultivars are to be bred and adequate production is to be sustained to boost the rural economies.

Research objectives

Given the foregoing, this study aimed at:

i. Appraising the farmer situation to obtain information on factors that might impact on dual- purpose cultivar production;

(24)

8 ii. Soliciting farmers and non-farmer stakeholders’ views and perceptions on dual-purpose

sorghum and the feasibility of their utilisation;

iii. Screening sorghum germplasm for grain yield potential and stem sugar traits;

iv. Investigating the gene action involved in the inheritance of grain yield potential and stem sugar in dual-purpose sorghum cultivars in selected tropical low and mid-altitude environments in Mozambique, South Africa and Zimbabwe;

v. Determining the levels of heterosis and cultivar stability for grain yield components and stem sugar traits in dual-purpose sorghum cultivars in selected tropical low and mid- altitude environments in Mozambique, South Africa and Zimbabwe;

vi. Investigating the relationship between grain yield potential and stem sugar traits in dual- purpose sorghum cultivars across selected tropical low and mid-altitude environments in Mozambique, South Africa and Zimbabwe; and

vii. Determining fertility restoration capabilities of introduced and regional sorghum germplasm used as male parents in hybrids formed from crosses with male sterile female lines from ICRISAT in selected southern African environments.

Research hypotheses

The research tested the following hypotheses:

i. Farmers and non-farmer stakeholders’ are aware of the dual-purpose sorghum cultivars and their potential benefits;

ii. There is high genetic diversity for grain yield potential and stem sugar traits among the germplasm collection held at the African Centre for Crop Improvement in South Africa;

iii. Grain yield potential and stem sugar traits in dual-purpose sorghum are controlled by genes that act predominantly in an additive manner;

iv. There are high levels of heterosis for grain yield and stem sugar traits that can be exploited to increase mean performance for the traits in dual-purpose sorghum hybrid cultivars;

v. Grain yield potential and stem sugar traits are independent of each other in dual-purpose sorghum cultivars; and

vi. Introduced and regional sorghum germplasm used as male in hybrid combination with male-sterile lines have the capacity to restore hybrid fertility in selected tropical low- and mid-altitude ecologies in South Africa.

(25)

9 Structure of the thesis

This thesis consists of eight distinct chapters in accordance with a number of activities related to the afore-mentioned objectives. Some overlap and repetition may exist between the chapters as they were written as independent journal papers containing all the necessary information, some of which might have been presented in other chapters. Some of the papers have already been published or accepted for publication and they are indicated in the thesis.

Chapter Title

- Introduction to thesis 1 A review of the literature

2 An appraisal of the factors impacting on crop productivity of small-scale farmers in the semi-arid environments in Zimbabwe and their implication for crop improvement goals and policy interventions

3 Development of sorghum for bio-energy: a view from stakeholders in Zimbabwe and South Africa

4 Variability for grain yield components and stem sugar traits for the development of dual-purpose sorghum cultivars for grain and bioenergy

5 Heterosis, combining ability and cultivar superiority of sorghum germplasm for stem- sugar traits across six environments

6 Combining ability, heterosis and cultivar superiority of sorghum germplasm for grain yield traits across tropical low and mid altitude environments

7 Relationship between grain yield components and stem sugar traits in dual-purpose sorghum germplasm

8 Fertility restoration capacities of southern African and introduced sorghum lines as measured by hybrid seed set across tropical low- and mid-altitude environments 9 An overview of the research findings

References

Anderson, I.C. 2005. Ethanol form sweet sorghum. Iowa Energy Centre, Iowa State University.

(26)

10 http://www.energy.iastate.edu/renewable/biomass/cs-anerobic2.html. Accessed 09 May 2006.

Baocheng, C., Qiaoying, L., Hong, J. and Lerong, L. 1986. An analysis of genetic effects of containing the sugar on different stems in sorghum. Acta Agronomica Sinica (China) 12:

39-42 (In Chinese).

Bello, D., Kadams, A.M., Simon, S.Y. and Mashi, D.S. 2007. Studies on genetic variability in cultivated sorghum (Sorghum bicolor L. Moench) cultivars of Adamawa State Nigeria.

American-Eurasian Journal of Agriculture and Environmental Science 2: 297-302.

Butler, J.L. 1992. Dedicated agricultural and herbaceous crops. Biomass and Bioenergy 2: 39- 44.

Chantereau, J. and Nicou, R. 1994. Sorghum. The Tropical Agriculturist Series. CTA Wageningen, Netherlands; Macmillan, London.

Chivasa, W., Chiduza, C., Nyamudeza, P. and Mashingaidze, A.B. 2001. Biodiversity on-farm in semi-arid agriculture: A case study from a smallholder farming system in Zimbabwe. The Zimbabwe Science News 34: 13-18.

Claassen, P.A.M., de Vrije, T., Budde, M.A.W., Koukios, E.G., Glynos, A. and Réczey, K. 2004.

Biological hydrogen production from sweet sorghum by thermophilic bacteria. 2^nd World conference on biomass for energy, industry and climate protection, 10-14 May 2004, Rome, Italy, pp 1522 – 1525.

Cruz, C.D. and Regazzi, A.J. 1994. Modelos biométricos aplicados ao melhoramento genético.

Viçosa, Imprensa Universitária, 394p. (English translation).

Derera, J., Tongoona, P., Langyintuo, A., Laing, M.D. and Vivek, B. 2006. Farmer perceptions on maize cultivars in the marginal eastern belt of Zimbabwe and their implications for breeding.

African Crop Science Journal 14:1-15.

Dixon, J., Gulliver, A., Gibbon, D. and Hall, M. 2001. Farming systems and poverty: Improving farmers’ livelihoods in a changing world. FAO and World Bank. Rome and Washington.

412pp.

Dolciotti, I., Mambelli, S., Grandi, S. and Venturi, G. 1998. Comparison of two sorghum genotypes for sugar and fibre production. Industrial Crop Products 7: 265-272.

FAO, 2002. Sweet sorghum in China. Agriculture 21 Magazine http://www.fao.org/ag/magazine/0202sp2.htm. Accessed 09 May 2006.

FAO, 2009. FAOSTAT. http://faostat.fao.org/site/567/DesktopDefault.aspx?PageID=567#ancor Accessed 09 November 2009.

(27)

11 Ferraris, R. and Charles-Edwards, D.A. 1986. A Comparative analysis of growth of sweet and forage sorghum crops. II accumulation of soluble carbohydrates and nitrogen. Australian Journal of Agricultural Research 37: 513-522.

Food Security Assessment Unit. 1998. Off-season production survey in southern Somalia January-March 1998 http://www.fsausomali.org/uploads/Other/274.pdf Accessed 16 February 2009.

Gnansounou, E., Dauriat, A. and Wyman, C.E. 2005. Refining sweet sorghum to ethanol and sugar. Economic trade-offs in the context of North China. Bioresource Technology 96:985- 1002.

Goyal, S.N. and Kumar, S. 1991. Combining ability for yield components and oil content in sesame. Indian Journal of Genetics and Plant Breeding 51: 311–314.

Guiying, L., Weibin, G., Hicks, A. and Chapman, K.R. 2000. A training manual for sweet

sorghum. FAO/CAAS/CAS, Bangkok, Thailand.

http://ecoport.org/ep?SearchType=earticleView&earticleId=172&page=2273 Accessed 15 November 2007.

Haussmann, B.I.G., Obilana, A.B., Ayiecho, P.O., Blum, A., Schipprack, W. and Geiger, H.H. 1999.

Quantitative-genetic parameters of sorghum [Sorghum bicolor (L.) Moench] grown in semi- arid areas of Kenya. Euphytica 105: 109-118.

Hugos, F., Makombe, G. Namara, R.E. and Awulachew, S.B. 2009. Importance of irrigated agriculture to the Ethiopian economy: capturing the direct benefits of irrigation. Colombo, Sri Lanka: International Water Management Institute. 37p. (IWMI Research Report.

Kenga, R., Alabi, S.O. and Gupta, S.C. 2004. Combining ability studies in tropical sorghum [Sorghum bicolor (L.) Moench]. Field Crops Research 88: 251-260.

Kudadjie, C.Y., Struik, P.C., Richards, P. and Offei, S.K. 2004. Assessing production constraints, management and use of sorghum diversity in north-east Ghana: a diagnostic study.

Netherlands Journal of Agricultural Science 52-3: 371-391.

Li, Y. and Li, C. 1998. Genetic Contribution of Chinese Landraces to the Development of Sorghum Hybrids. Euphytica 102: 47-55.

Makanda, I., Tongoona, P., Madamba, R., Icishahayo, D. and Derera, J. 2009. Path-coefficient analysis of bambara groundnut pod yield components at four planting dates. Research Journal of Agriculture and Biological Sciences 5: 287-292.

Mangisoni, J.H., Kachule, R., Kalinda, T., Chilongo, T., Simfukwe, M. and Tostão, E. 2007. Input voucher study in Malawi, Mozambique and Zambia: Final Synthesis Report. Food